# Welt Python SDK 0.13.1

> Python SDK for Welt: fit qualified tabular models on DataFrames, reuse predictors, forecast, load research datasets and run native causal discovery. Each link is this release's Markdown source.

For the hosted MCP server, the CLI and product-level agent docs, see https://welt.ergodic.dev/llms.txt

For local identification and effect estimation on a Welt graph, see ergodic (`pip install "ergodic[welt]"`): https://docs.ergodic.dev/llms.txt

## Get started

- [Your first prediction](https://ergodic-ai.github.io/welt-python/v0.13.1/start.md)
- [Connect your workspace](https://ergodic-ai.github.io/welt-python/v0.13.1/connect.md)
- [Install to your first real request](https://ergodic-ai.github.io/welt-python/v0.13.1/onboarding.md)
- [Learn in a notebook](https://ergodic-ai.github.io/welt-python/v0.13.1/notebooks.md)

## Guides

- [Predict with your own table](https://ergodic-ai.github.io/welt-python/v0.13.1/own-data.md)
- [Predict a number](https://ergodic-ai.github.io/welt-python/v0.13.1/regression.md)
- [Reuse a fitted predictor](https://ergodic-ai.github.io/welt-python/v0.13.1/reuse.md)
- [Selecting and swapping foundation models](https://ergodic-ai.github.io/welt-python/v0.13.1/model-selection.md)
- [Compare models with cross-validation](https://ergodic-ai.github.io/welt-python/v0.13.1/evaluations.md)
- [Forecast a univariate time series](https://ergodic-ai.github.io/welt-python/v0.13.1/forecasting.md)
- [Hidden model runs and univariate forecasting](https://ergodic-ai.github.io/welt-python/v0.13.1/model-runs.md)
- [Research datasets](https://ergodic-ai.github.io/welt-python/v0.13.1/research-datasets.md)
- [Wait and recover](https://ergodic-ai.github.io/welt-python/v0.13.1/async-and-errors.md)
- [Causal discovery](https://ergodic-ai.github.io/welt-python/v0.13.1/causal-discovery.md)
- [SDK guide](https://ergodic-ai.github.io/welt-python/v0.13.1/usage.md)

## Reference

- [Python reference](https://ergodic-ai.github.io/welt-python/v0.13.1/reference.md)
- [Implemented and planned capabilities](https://ergodic-ai.github.io/welt-python/v0.13.1/capabilities.md)
- [Welt CLI for agents](https://ergodic-ai.github.io/welt-python/v0.13.1/cli.md)

## Releases

- [Upgrade your client](https://ergodic-ai.github.io/welt-python/v0.13.1/migration.md)